Optical coherence photoacoustic microscopy for 3D cancer model imaging with AI-assisted organoid analysis

Cancer organoids and cancer spheroids are 3D cell culture models with distinct yet overlapping purposes in cancer research. Various commercially available optical imaging techniques have been employed to study these cell cultures, but these methods suffer from various limitations such as the requirement of fluorescence labeling, complicated sample handling, and limited image volume size. In this work, we demonstrate a multimodal optical coherence photoacoustic microscopy (OC-PAM) system for the study of these models, overcoming these limitations. We first performed a longitudinal study using optical coherence microscopy (OCM) for breast cancer organoids. Using the OCM imaging results, artificial intelligence (AI)-based algorithms were developed to automatically segment individual organoids and classify their viability over time using a radiomics texture feature approach, enabling robust, quantitative tracking and classification at the single-organoid level. To supplement OCM’s contrast, we then performed OC-PAM imaging of spheroid models with both melanin positive and melanin negative cells. In the second study, the OC-PAM images clearly mapped the distribution of melanin positive cells hidden amongst melanin negative cells. These results suggest that OC-PAM coupled with AI techniques can be a powerful tool to study cancer organoids and cancer spheroids.